DISCUSSION PAPER
N o ve m b e r 2 0 0 8
„
RFF DP 08-41
Opportunities and
Threats in Targeting
Antimalarials for
the AMFm
The Role of Diagnostics
Christopher J.M. Whitty, Heidi Hopkins, Evelyn
Ansah, Toby Leslie, and Hugh Reyburn
Prepared for the Consultative Forum on AMFm—the Affordable
Medicines Facility-malaria
September 27–28, 2008
Wash in g ton , DC
1616 P St. NW
Washington, DC 20036
202-328-5000 www.rff.org
Opportunities and Threats in Targeting Antimalarials for the
AMFm: The Role of Diagnostics
Christopher J.M. Whitty, Heidi Hopkins, Evelyn Ansah, Toby Leslie, and Hugh Reyburn
Abstract
In malaria-endemic countries, people commonly assume they have malaria when sick and treat
themselves accordingly. The Affordable Medicines Facility-malaria (AMFm) will make effective drugs
more available everywhere. If longstanding problems can be successfully addressed, the use of
microscopy and rapid diagnostic tests (RDTs) for malaria diagnosis alongside AMFm (though not
necessarily directly part of it) could improve the management of both malaria and other febrile illness, as
well as the cost-effectiveness of AMFm. In peripheral areas, RDTs are the only practical option, but
available RDTs have limitations: all-or-none test results, variable heat stability, an inability to diagnose
non-falciparum malaria and safety risks (especially HIV and hepatitis B) related to blood sampling. Of
equal concern is that negative test results—meaning no malaria—are often ignored and patients treated
anyway. R&D to solve technical problems and operational research on better ways to deploy RDTs and to
make diagnosis count are needed.
Key Words: malaria, diagnosis, RDT, microscopy, cost-effectiveness, Africa
JEL Classification Numbers: I10 and I18
© 2008 Resources for the Future. All rights reserved. No portion of this paper may be reproduced without
permission of the authors.
Discussion papers are research materials circulated by their authors for purposes of information and discussion.
They have not necessarily undergone formal peer review.
Contents
Executive Summary ................................................................................................................. i
Introduction............................................................................................................................. 1
Diagnostic Tests for Malaria.................................................................................................. 3
Microscopy ......................................................................................................................... 4
Rapid Diagnostic Tests ....................................................................................................... 6
What Is Relevant for AMFm?............................................................................................. 8
How Diagnostic Tests Are Used in Africa and Asia............................................................. 9
Do RDTs Change Prescribing Behavior? ........................................................................... 11
What Is Relevant for AMFm?........................................................................................... 13
The Situation in Asia ............................................................................................................ 13
Conclusions............................................................................................................................ 15
References.............................................................................................................................. 19
Resources for the Future
Whitty et al.
Opportunities and Threats in Targeting Antimalarials for the
AMFm: The Role of Diagnostics
Christopher J.M. Whitty, Heidi Hopkins, Evelyn Ansah, Toby Leslie,
and Hugh Reyburn∗
Executive Summary
Historically and today, a large proportion of patients with febrile illness in places where
malaria is common, especially in Africa, are treated with antimalarial drugs, but without specific
diagnosis. The exception to the lack of diagnosis has been blood smear slides examined under
the microscope for malaria parasites, a practice limited largely to hospitals and formal clinics.
More recently, several rapid diagnostic tests (RDTs) for malaria have become available and the
question of how they can best be deployed—and specifically, whether they should be linked to
the Affordable Medicines Facility-malaria (AMFm)—arises. Targeting antimalarials to those
who have malaria and identifying and treating other causes of serious febrile diseases is an
undisputed long-term goal, but is far from the current state of affairs.
In rural villages in Africa, where health-care institutions are rare, chloroquine and
sulfadoxine-pyrimethamine (SP) have been the main malaria drugs available, offered by shops
and street vendors and purchased for patients for empirical use. Replacing those drugs with
ACTs is the purpose of AMFm and a necessary component of malaria control. ACTs are
relatively safe and very effective. A person with malaria is likely to be helped, and a person
without malaria is not likely to be hurt (even if treating the uninfected person is not desirable).
In the peripheral, nonformal, and private sectors in Africa, where AMFm is likely to have
its greatest impact, RDTs are the only realistic option. Microscopy, to be cost-effective, depends
on high throughput, which is likely only in the formal sector. But are currently available RDTs
ready for widespread deployment? RDTs have a number of technical limitations, including an
all-or-none test result, variable heat stability, and in diagnosing non-falciparum malaria.
∗
Christopher J.M. Whitty, London School of Hygiene & Tropical Medicine, London, UK; Heidi Hopkins,
FIND,Uganda; Evelyn Ansah, Dangme West, Ghana; Toby Leslie, London School of Hygiene & Tropical Medicine,
London, UK, and Health Protection and Research Organisation, Afghanistan; and Hugh Reyburn, Joint Malaria
Programme, Tanzania. This paper was commissioned for Resources for the Future's Consultative Forum on the
Affordable Medicines Facility-malaria (AMFm) under a grant from the Bill & Melinda Gates Foundation.
i
Resources for the Future
Whitty et al.
Safety needs to be taken seriously. There is a theoretical risk of RDTs leading to true
cases of malaria not being treated, and of deploying a technology that requires blood sampling by
relatively untrained users in areas of high HIV and hepatitis B prevalence.
In the formal health-care sector in Africa, health-care workers are responsive to positive
malaria tests, using either microscopy or RDTs, prescribing antimalarials for virtually everyone
who tests positive. But about half the time, a negative test elicits the same prescription. Data
from Asia, though limited, reveal a similar picture. The reasons clinicians respond irrationally to
diagnostic tests are complex and may be difficult or slow to change. Data from outside the
formal sector are too scant to be informative.
The bottom line is that these problems are recognized and will exist with or without
AMFm. Improving diagnosis and the use of diagnostic tests alongside AMFm (though not
necessarily a direct part of it) has the potential to improve management of both malaria and other
febrile illness as well as the cost-effectiveness and sustainability of AMFm. The impact of
improved targeting is greatest in areas where the incidence of malaria is lower and other causes
of febrile disease are relatively more important. R&D to solve some of the technical problems
and operational research to develop better ways to deploy RDTs and to make diagnosis count are
also needed.
ii
Resources for the Future
Whitty et al.
Opportunities and Threats in Targeting Antimalarials for the
AMFm: The Role of Diagnostics
Christopher J.M. Whitty, Heidi Hopkins, Evelyn Ansah, Toby Leslie,
and Hugh Reyburn∗
Introduction
The opportunity to expand the use of effective antimalarials, in practice artemisinin
combination therapies (ACTs), with the AMFm initiative is potentially a major step forward and
undoubtedly very welcome, both within and outside the public sector. There is, however, a
potential downside to expanding use in areas where diagnostic facilities are weak: increasing the
number of febrile patients treated with an antimalarial may not translate into as large an increase
in those who have malaria being treated with an antimalarial. There are many causes of febrile
illness, and it is essential that antimalarial therapy is targeted as much as possible toward those
who have the disease. This is important for several reasons.
First, important alternative diagnoses may be missed. Evidence is strong from many parts
of Africa as well as outside it that the high burden of mortality in childhood is attributable to
many treatable febrile illnesses; malaria is certainly not the only cause of death and in many
settings is not even the most important (Olivar et al. 1991; Sowunmi and Akindele 1993;
Ndyomugyenyi et al. 2007). Treating meningitis or pneumonia with an antimalarial, for example,
is at best useless and possibly, if it leads to delays in diagnosis and appropriate treatment, even
hazardous (Berkley et al. 1999; Reyburn et al. 2004; Berkley et al. 2005; Kallander et al. 2008;
Orimadegun et al. 2008). The alternative to improved targeting of treatment for nonsevere febrile
illness without diagnosis is syndromic management with antibiotics and antimalarials, but though
this approach may have some attractions, it has very limited in-country support, with concerns
about antibiotic resistance widely voiced.
Second, today’s ACTs generally appear safe and well tolerated, but no drug is without
risk (Maiteki-Sebuguzi et al. 2008). Treating malaria with an antimalarial is a balance of the
∗
Christopher J.M. Whitty, London School of Hygiene & Tropical Medicine, London, UK; Heidi Hopkins,
FIND,Uganda; Evelyn Ansah, Dangme West, Ghana; Toby Leslie, London School of Hygiene & Tropical Medicine,
London, UK, and Health Protection and Research Organisation, Afghanistan; and Hugh Reyburn, Joint Malaria
Programme, Tanzania. This paper was commissioned for Resources for the Future's Consultative Forum on the
Affordable Medicines Facility-malaria (AMFm) under a grant from the Bill & Melinda Gates Foundation.
1
Resources for the Future
Whitty et al.
relatively small risk of the drug against the very high risk of the disease. If, however, the great
majority of those treated with an antimalarial do not have the disease and therefore cannot
benefit, the risk-benefit changes presenting a significant safety issue for drugs, and especially
newer drugs.
Third, the cost-effectiveness of any antimalarial program will be substantially reduced if
most of the money spent on antimalarials goes to treat people who do not have malaria (Snow et
al. 2003; Yeung et al. 2008). This is potentially an important limitation on the long-term
sustainability of funding, and, in particular, of maintaining widespread political support. Because
the success of upscaled malaria interventions (i.e., access to effective drugs and insecticidetreated nets) will result in reduced incidence of disease, without diagnosis, cost-effectiveness will
only decrease; the proportion of febrile cases with malaria decreases though there may be no
great change in the number of cases administered antimalarials. Being able to shift resources
from drugs to other interventions as this change occurs will make diagnosis more important over
time.
Fourth, the risk of increasing the potential for the emergence and particularly the spread
of antimalarial drug-resistance, especially to the partner drugs in ACTs, and to a lesser extent,
the artemisinins themselves, is small but real, dependent on endemicity. Low levels of drug
circulating in many individuals in a population is an ideal situation to speed up the spread of
resistance (Payne 1988; Talisuna et al. 2004). Recent data suggest that artemisinin-tolerant
parasites may exist on the Thai-Cambodian border, which the WHO recognizes as a potential
global emergency with serious implications for malaria control efforts (WHO 2008). Strategies
to delay the spread of resistance and preserve ACTs as effective antimalarial therapies are critical
to long-term malaria control goals (White and Olliaro 1996; Duong et al. 2004).
Fifth, if a large proportion of those treated with an antimalarial do not get better because
they are not actually suffering from malaria, this can undermine popular support for new and
effective antimalarials. Additionally, real gains (and losses) in malaria control programs can be
missed when all fever cases are classified as malaria.
It is thus essential to move toward targeting antimalarials so that they go to those with
malaria and not those whose fever is caused by a different infection. AMFm, by moving ACTs
farther into the community, beyond the formal health-care system, could be a vehicle for better
diagnostic practices—thus improving management of all causes of febrile illness. If diagnostic
issues are ignored, AMFm could simply increase the amount of ACTs given to patients who do
not have malaria. The diagnostic issues are certainly not a reason to delay or stop AMFm;
2
Resources for the Future
Whitty et al.
increasing the use of effective antimalarials in peripheral settings, where the poorest live, is
essential. Technical decisions about AMFm must, however, take diagnostics and drug targeting
into account if AMFm’s potential is to be maximized.
There is now increasing evidence of a reduction in malaria transmission in many of the
countries where malaria is a major problem (Barnes et al. 2005; Nyarango et al. 2006; Bhattarai
et al. 2007; Okiro et al. 2007; Guerra et al. 2008; Sievers et al. 2008). This phenomenon is
extremely welcome, and the use of artemisinins may well be contributing to the decrease, in
some cases substantially. This means, however, that antimalarial targeting will become more
rather than less important. A diagnostic system in which most febrile illness is treated with
antimalarials is rational when most children with potentially fatal febrile illnesses have malaria.
This approach is harder to justify when only a minority do. In addition when incidence
decreases, the age distribution of cases widens. This observation highlights a technical limitation
with the current RBM Monitoring and Evaluation Reference Group (MERG) indicators, which
emphasize the need for children with febrile illnesses—rather than febrile illness with
parasitemia—to be treated with an effective antimalarial within 24 hours. This empiric approach
makes sense in many settings but is less appropriate when malaria is a relatively smaller problem
than bacterial disease. This is the situation for much of highland and urban Africa and
increasingly for some previously highly endemic parts of rural Africa as well (Barnes et al. 2005;
Nyarango et al. 2006; Bhattarai et al. 2007; Ndyomugyenyi et al. 2007; Okiro et al. 2007; Sievers
et al. 2008).
The rest of this paper summarizes the current knowledge about the diagnosis of malaria
in the field, reviews technical aspects of malaria diagnostic tests, looks at the potential costeffectiveness of these tests, and then examines the experience so far in trying to improve
diagnostic practices within the formal health-care sector. It finishes with discussion of the
potential for extending diagnostic services into the nonformal sector in Africa, which is the
where AMFm is likely to have its greatest impact, and the possible role of AMFm in facilitating
greater use of rapid diagnostic tests (RDTs). The discussion is somewhat speculative because
data on this topic are scarce, making it a target for operational research and evaluation of
different strategies.
Diagnostic Tests for Malaria
It is widely accepted that clinical algorithms, which use only symptoms to identify those
with malaria, are difficult to use and have low sensitivity and specificity (Redd et al. 1996;
Luxemburger et al. 1998; Tarimo et al. 2001; Chandramohan et al. 2002). The options therefore
3
Resources for the Future
Whitty et al.
are either to treat all individuals with a fever or recent history of fever, or to use a diagnostic test.
For malaria, the choice of test is between microscopy, which identifies parasites directly in
blood, and RDTs, which are immunochromatographic tests that detect malaria antigens in the
blood and can be read visually with the naked eye. To summarize, light microscopy, provided it
can be done well, remains the gold standard in settings where many patients are to be tested, but
in more peripheral settings in Africa is unlikely to be either feasible or cost effective (although
use of portable microscopy has been used successfully in parts of Asia). Therefore, if diagnostic
tests are to be considered for use in these peripheral settings most relevant for AMFm, the choice
is likely to be between RDTs and clinical diagnosis based solely on fever. Important differences
between microscopy and the most prominent types of RDT are summarized in table 1.
Other approaches, such as using the polymerase chain reaction (PCR) and immunological
tests for malaria are also available but are useful only for research in field contexts in Africa,
Asia, and South America. There is no realistic probability of PCR becoming a useful clinical tool
in low-resource health-care settings in the foreseeable future. Immunological tests, even when
well performed, are unreliable for acute diagnosis of malaria, though they are useful in a number
of other contexts, such as blood banks. We therefore concentrate on the two standard
technologies of microscopy and RDTs and discuss possible developments to these within a fiveto-ten year timeframe.
Microscopy
The mainstay of malaria diagnosis for the last 100 years has been light microscopy of
blood stained with Field and Giemsa variations. A more recent development is fluorescence
microscopy (QBC II), which reportedly improves sensitivity and user effectiveness (Gray et al.
1991; Gay et al. 1996; Guy et al. 2007). The vogue for fluorescent technology seems to have
largely passed though fluorescence microscopes are still used in some parts, particularly in East
Africa where many were donated, but those that break tend not to be replaced and, in practice,
they have had relatively little additional impact over light microscopy.
It is easy to write off light microscopy as yesterday’s technology. That it has survived so
long is largely because it is still the best method, provided there are good microscopes, good
slides and stains, and highly trained and motivated technicians with enough time. In expert
hands, the lower limit of detection of malaria in thick and thin films is about 50 parasites/µl of
blood (assuming a total RBC count of 5 x 106/µl of blood), which is equivalent to 0.001% of
RBC infected parasites per micro-liter (Moody 2002). The detection of low-level parasitemias
may not have a great bearing on diagnosis in most African settings, given that high (and
4
Resources for the Future
Whitty et al.
therefore relatively easily detected) parasitemia is the norm in clinical cases. It is, however,
important in areas of low transmission and in pregnant women.
The microscopes used for diagnosing malaria have remained essentially unchanged for
the last 50 years because they do their job effectively and are generally robust. They can be
adapted to situations where there is no power, using sunlight reflected in mirrors as a light
source, though sensitivity tends to decrease compared with powered light sources. A number of
portable microscopes have been developed that are effective in certain situations. It is therefore
premature to write an obituary for light microscopy, even though we do not anticipate any new
technological developments. It suffers, however, from a number of drawbacks:
•
The initial capital outlay is significant and the cost of employing trained technicians and
supplying them with adequate reagents and equipment is not trivial. This cost is justified
in high throughput settings, where the cost-effectiveness of microscopy tends to dominate
that of any other diagnostic method (Jonkman et al. 1995; Lubell et al. 2007). Light
microscopy is therefore likely to remain the ideal standard in hospital outpatient and
other high-volume settings for the foreseeable future. The good examples of National
Malaria programs that were successful using microscopes in the periphery, such as
Thailand and Vietnam, seem unlikely to be replicated in Africa. Microscopy’s costeffectiveness decreases significantly as the throughput of true cases of malaria decreases,
such as in clinics in the periphery, in pharmacist shops, and in community settings.
Certain externalities, however, such as using the same staff and equipment for TB or
stool tests, may increase cost-effectiveness in lower throughput settings.
•
Providing new microscopes has proved relatively easy for donors and ministries of
health, but maintaining them is far more challenging. Microscopes are relatively simple
mechanically, but problems with fungal hyphae and physical damage (to lenses, focusing
mechanisms, and spare parts such as bulbs) mean that many microscopes are either
unusable or usable only at standards far below optimal.
The quality of slides and stains is often poor and frequent power outages are a significant
problem when generators are not available.
•
•
Training and maintaining staff morale are ongoing issues. Judged by a gold-standard of
double-read research slides, many studies have demonstrated that sensitivity of
microscopy in operational practice, even with recently replaced equipment, often falls
below 70 percent and that specificity is equally poor (Durrheim et al. 1997; Stow et al.
1999; Coleman et al. 2002). Clinicians are aware of this deficiency, which is probably
largely the reason for a widespread belief in African and Asian settings that “slidenegative malaria” is a common problem. Series from Europe, the United States and
Southeast Asia, where resource constraints are fewer and equipment is of better quality,
demonstrate that slide-negative malaria is in fact exceptionally rare, even in settings
where most people are not immune and very low parasitemia can thus cause disease.
5
Resources for the Future
Whitty et al.
Much more often, malaria is called slide-negative when the microscopy result is in fact
false-negative because of poor slide preparation or examination technique. A negative
cycle therefore tends to occur when clinicians ignore slide results, slide readers learn that
clinicians ignore their results, slide reader motivation drops still further, and diagnostic
accuracy decreases even more. This has mainly to do with the attitude of clinicians and
may not be true at other levels of care.
Various attempts have been made to improve microscopy training and equipment
(Ssekabira submitted for publication). The general experience is that such efforts begin well but
deteriorate rapidly. Maintaining an excellent microscopy service in high throughput settings is
therefore an ideal, but often unrealized.
Rapid Diagnostic Tests
Rapid diagnostic tests (RDTs), commercially available as cassettes or dipsticks, are
increasingly common and have been shown effective in a variety of clinical settings. RDTs are
performed by placing a drop of blood on a cassette or dipstick and applying a buffer solution to
diffuse the blood across a membrane, which then shows control and test lines, similar to rapid
HIV tests or pregnancy tests. RDTs detect parasite antigens (proteins) in whole blood. RDTs
today are designed to detect one or more of three target antigens: histidine-rich protein 2 (HRP2),
parasite lactate dehydrogenase (pLDH), and aldolase. The majority of RDTs on the market detect
either HRP2 or LDH. More than 80 RDTs, mainly of these two types, are currently on the
market, (WHO/WPRO 2008) with a relatively rapid turnover of manufacturers.
Currently available RDTs have a number of advantages and disadvantages compared
with conventional microscopy. Several advantages are clear cut:
• RDTs are relatively simple to use, requiring minimal training to master the mechanics of
test preparation and interpretation (Premji et al. 1994; Mayxay et al. 2004; Rennie et al.
2007; Harvey et al. 2008).
•
RDT sensitivity and specificity in detecting parasitemia is comparable to good-quality
microscopy (Craig et al. 2002; Guthmann et al. 2002; Moody 2002; Hopkins et al. 2007)
and therefore better than an underskilled microscopist or a microscopist using poor
equipment.
•
In low throughput settings, when only a few patients who might have malaria are seen per
day RDTs are more cost-effective than microscopy(Bualombai et al. 2003; Shillcutt et al.
2008).
6
Resources for the Future
Whitty et al.
•
RDTs do not require electricity or special laboratory equipment.
•
RDTs are portable.
•
RDT are good for quality control because they can be verified days after being
performed.
•
Results of RDTs are understandable by the patients so that, in principle, they are less
likely to be ignored by the prescribers, although evidence for this in practice is lacking.
These last two advantages make them very useful for rapid screening, use in refugee
settings and, of particular importance for the current discussions, use in remote or rural settings.
RDTs do have a number of limitations as well:
•
Current tests are all-or-none, that is, results are either positive or negative, and do not
allow for quantification of parasitemia. This is not a limitation in settings where people
are not immune because any degree of parasitemia may signify a clinically significant
infection. It is more of a problem where transmission is high (malaria is meso- to holoendemic), and many children and adults (20 percent of the population might be typical)
carry parasites at any given time, but remain asymptomatic and essentially well (these are
not seen in the dispensaries). If they present with another cause of febrile illness, they
may well be correctly diagnosed as having malaria parasitemia, but the diagnosis may be
misleading about the real cause of their illness if the parasite count is very low.
Microscopy allows a quantitative assessment of parasitemia, which is more useful to
clinicians in this setting (although this advantage is lost where clinicians do not trust the
results of the microscopist). A patient with a low parasite count will still benefit from
malaria treatment, but other potentially serious diseases may be left undiagnosed.
•
Current tests are not heat stable (Jorgensen et al. 2006; Chiodini et al. 2007). This varies
by test type, but all RDTs are at risk of deterioration and reduced sensitivity when they
are exposed to heat or humidity for prolonged periods, some becoming essentially useless
over the course of just a few days when exposed to temperatures above 40° C. RDTs are
commonly subject to high temperatures between leaving the factory gate and use in the
field. Therefore, unless they can be deployed using a cold chain or made substantially
more heat stable, they are poorly suited for use in Africa and other tropical areas where
malaria is a serious issue. This is less of a problem for heavy users (such as medical
NGOs) because they are used quickly in clinics and the NGOs are aware of temperature
vulnerabilities.
•
In high throughput settings, the cost of RDTs, most of which retail at around 60 cents per
test, is substantially higher than light microscopy and therefore less cost effective than
microscopy (assuming a level of accuracy of microscopy).
7
Resources for the Future
Whitty et al.
•
Sensitivity and specificity in placental malaria in pregnancy has not been determined
(Mockenhaupt et al. 2006; Uneke 2008). Very sensitive tests, i.e., tests that detect very
nearly all cases (accepting a higher level of false positives), are needed during pregnancy.
•
HRP2-based RDTs may show positive results for days to weeks after an effectively
treated episode of malaria because of persistent circulating antigen. This reduces their
specificity in settings where malaria is common and could lead to other febrile illnesses
being misdiagnosed as malaria on the basis of a genuinely positive test but a falsepositive one in clinical practice (B. Nadjm, personal communication) (Mayxay et al.
2001; Tjitra et al. 2001; Singh and Shukla 2002; Swarthout et al. 2007).
Three of these disadvantages may be overcome by technological advances. Heat stability
has the potential to improve, and attempts by WHO to provide a systematic lot-testing system is
a financial incentive for companies to maintain their manufacturing standards beyond the period
when the tests are being assessed for initial deployment (http://www.wpro.who.int/sites/rdt).
Second, it seems likely that quantitative tests will be deployed in due course, which will allow
differentiation between high and low parasite counts, but this is still some years away. Such
technological advances are unlikely to reduce costs and may well increase them. Third, it seems
probable that, as the technology matures, costs will decrease over time, but it is improbable that
in high transmission, high throughput settings RDTs will ever exceed the cost-effectiveness of
accurate microscopy.
Without doubt, however, the existence of RDTs has revolutionized the possibilities for
parasite-based diagnosis beyond the hospital (or even clinic) setting. They may even have a role
in hospitals and other higher-level facilities where light microscopy is not reliable or during offhours.
One fact frequently forgotten as a practical issue is that RDTs still require blood to be
drawn. The safety of deploying lancets for repeated fingerpricks in nonformal health-care
settings where HIV prevalence is high (much of Africa, particularly Southern Africa) or hepatitis
B is common (also Africa, particularly West Africa, and many parts of Asia) is an important
consideration in discussions of RDT deployment outside the formal health-care sector.
Noninvasive diagnostic tests, such as using urine or saliva, have been proposed but are not
currently near deployment. They may never reach that stage and plans should not be made on the
assumption that they will.
What Is Relevant for AMFm?
If AMFm is approved, the ACT expansion it will lead to will occur in the peripheral,
informal, and private sector (meaning small unregulated shops). Microscopy will, for reasons
8
Resources for the Future
Whitty et al.
outlined, continue to play little role in this sector in Africa. The choice will therefore be between
continuing with clinical diagnosis (as with chloroquine), and deploying RDTs as well as the
ACTs. Here we outline behavioral issues on how diagnostic tests are used in clinical practice.
First, though, we need to answer three questions about the technology.
•
Could deploying RDTs lead to a reduction in correct treatment of true malaria cases
because of false-negative test results? This depends, critically, on consistent
manufacturing quality, test robustness under typical transport and storage conditions, and
correct test performance by end users. In principle, syndromic treatment should be almost
100 percent sensitive but very nonspecific. In practice, however, it is as often applied
unsystematically. The 100 percent sensitivity is thus theoretical, but diagnostic tests may
improve its specificity.
•
Can deploying RDTs be made safe from the risk of increasing transmission of
bloodborne infections, especially HIV and hepatitis B? This will depend on adequate
training in safe blood taking, a free supply of gloves and lancets (eliminating the
incentive to use either twice), and a reliable method for safe disposal of sharps.
•
Is there a financial incentive to shopkeepers and others to use RDTs, if they are available,
and, if not, can incentives be devised if it is thought desirable?
If the answer to the first question is yes, or the answer to the second or the third is no, the
deployment of rapid tests alongside AMFm is unlikely to be advisable even if the considerable
behavioral challenges outlined can be addressed.
How Diagnostic Tests Are Used in Africa and Asia
It is well recognized that the majority of children and pregnant women with malaria do
not access formal health care, (Breman 2001) and typically are either not treated with an
antimalarial or treated with an ineffective one. It is the second problem that the AMFm seeks to
address. The probability is high that before children and adults receive formal health care they
will not—despite having malaria—receive an antimalarial. The probability is equally high,
however, that once they do receive the care, they will receive an antimalarial regardless.
The majority of the information we have on diagnostic targeting of antimalarials to
people with malaria comes from the formal public sector (hospitals and clinics), rather than
peripheral dispensaries and the various elements of the private and informal sector where the
AMFm is likely to have greatest impact. Some degree of extrapolation is therefore necessary.
Extrapolation, however, carries risks: behavior in one segment of the health-care system (broadly
defined) may be very different from others. With this caveat in mind, a number of studies in
9
Resources for the Future
Whitty et al.
Africa have looked at variations on the question of what proportion of children and adults given
an antimalarial has actually had malaria. The answer has been consistent: overtreatment of
children and especially adults with antimalarials is substantial. Overdiagnosis is defined as
giving an antimalarial to a patient who does not have any malaria parasites detected on a malaria
test the clinician has requested. This is true for severe disease as well as for nonsevere disease,
but this review concentrates on nonsevere disease because it is most relevant to AMFm.
The picture that emerges from the literature across Africa is clear: clinicians almost
invariably respond to positive malaria tests by prescribing antimalarials, but often respond to
negative tests by ignoring them and prescribing antimalarials anyway. When diagnostic facilities
are available, half or more of those with negative test results are still treated for malaria (Hamer
et al. 2007; Reyburn et al. 2007; Zurovac et al. 2008). When diagnostic facilities are not present,
the proportion is even higher. Where malaria is common, most of those given an antimalarial
actually have malaria parasites but a significant minority does not. In the many settings when
malaria is not the predominant cause of febrile illness, that is, fewer than 10 percent of children
presenting with fever have malaria parasites, the proportion of negative tests treated stays the
same but the absolute numbers of those with a negative test treated with an antimalarial increase
substantially. In low transmission settings more than 90 percent, and in very low transmission
settings more than 99 percent, of children treated with an antimalarial do not have malaria,
despite tests being available (Ndyomugyenyi et al. 2007).
There is little doubt that this situation has not been helped by the ambivalent and
potentially confusing message from WHO, and subsequently from national malaria control
programs, that it is important to perform diagnostic tests for malaria in children under five in
high transmission settings, and that if the clinician thinks a child has malaria, the child should be
treated for malaria regardless of the test result (WHO 2006). This has also been interpreted as a
need for blanket treatment of malaria in all children with fever. It must be noted that the term
high transmission is not further defined in these guidelines. In true high transmission settings,
when the majority of febrile illness in young children is malaria, this approach is safe, provided
antibiotics are also considered. However, empiric treatment becomes problematic when this
logic, which is explicitly meant only for high-transmission settings, is applied across the
continent regardless of malaria incidence, almost all clinicians supposing their region to be a
high malaria transmission area. Nursing and medical curricula may contribute to this impression.
The true epidemiological situation in Africa is of course much more complex than this. In many
areas, such as highland and some urban areas, malaria transmission is low or even nonexistent
10
Resources for the Future
Whitty et al.
and has been so since records began. In other areas, transmission appears to be dropping
significantly. The issue, then, is far from trivial.
Recent data from African settings without facilities for diagnosing malaria are currently
sparse, but unsurprisingly find the same pattern of overdiagnosis (Olivar et al. 1991; Sowunmi
and Akindele 1993). Inevitably, however, evidence of misdiagnosis in both directions is also
greater, with true cases of malaria being missed. Because malaria symptoms are nonspecific,
especially in the early stages, only a diagnostic strategy that treated every unwell child as for
malaria could be guaranteed to miss no cases.
Remarkably few equivalent data from the formal health-care sector in Asia are available,
possibly in part because of an assumption that, because malaria is much less prevalent, it is much
less likely that it is overdiagnosed. Such an assumption may well be false. A recent study in
India, in an area with a relatively low endemicity for malaria, for example, demonstrated
substantial overdiagnosis. This may well not be typical in other areas of Asia, and especially
Southeast Asia, where experience from Thailand suggests RDTs have played a very useful role.
Because the population of India is greater than that of malaria-endemic Africa, this finding is not
trivial if it is consistent with practice elsewhere in India. Our empirical observations suggest that
the same may well be true in Pakistan and Afghanistan at least, but we are aware of no published
or unpublished data to support or refute this.
In summary, evidence of overdiagnosis of malaria in the formal health-care sector in
many parts of Africa is clear, and some evidence from South Asia suggests a problem there as
well, although this may well not be representative. Even when clinicians have tests and choose to
use them, they often ignore the results. We are not aware of comparable data from Central or
South America. The cost-effectiveness of RDTs erodes rapidly if negative tests for malaria are
ignored (Lubell et al. 2008; Lubell et al. 2008).
Do RDTs Change Prescribing Behavior?
The purpose of a test is to change diagnostic and treatment practice. With RDTs and
microscopy, we now have tools that provide accurate and relatively rapid results to guide case
management. Do prescribers respond to the provision of RDTs by changing diagnostic decision
making? The answer from a number of recent trials in the formal sector in Africa would appear
to be either no or not very much.
Two studies from Africa, one a randomized control trial and the other a major
observational study, have recently been published (Hamer et al. 2007; Reyburn et al. 2007). We
11
Resources for the Future
Whitty et al.
are aware of at least two other trials for which abstracts are available. These data are backed up
by a number of observational studies (Chandler et al. 2008; Zurovac et al. 2008). All seem to
suggest, first, that clinicians often fail to request a diagnostic test when it is clinically
appropriate, and, second, that when a test result is obtained, clinicians ignore both microscopy
results and RDTs.
In the two published studies, which were conducted in several epidemiological settings
(one in Tanzania and the other in Zambia), approximately 99 percent of patients with a positive
diagnostic test (either microscopy or RDT) and approximately 50 percent of those with a
negative RDT were prescribed an antimalarial. When only 1 percent of all febrile illness is
malaria, the great majority (more than 90 percent) of all antimalarials prescribed were given to
patients for whom the clinician had chosen to undertake a test, had received a negative result,
and had prescribed an antimalarial regardless. The impact of negative RDTs on increasing
antibiotic prescribing was slightly greater but certainly not startling. Comparing microscopy with
RDTs in this randomized trial demonstrated that the two tests were treated almost exactly the
same, with no greater credence given to RDT results.
Of the studies reported at meetings in abstract form only, an observational study in
Zambia found a similar pattern, though negative RDT results were respected more often than
negative microscopy results.
All the studies reported here provided a half-day or one-day training package, designed to
be realistic in terms of what a national malaria control program could deliver across a country,
focusing on details of RDT performance, and typically with a brief review of national case
management guidelines. The results of more intensive training efforts on health worker
prescribing behavior in Uganda are encouraging. A one-week integrated course for health
workers at facilities with microscopy significantly decreased unnecessary ACT prescriptions
(Ssekabira submitted for publication). In addition, preliminary data from an evaluation of an
RDT training course targeted to workers at peripheral health facilities, which included clear
guidelines on management of patients with positive and negative RDT results, dramatically
decreased unnecessary antimalarial prescriptions while maintaining satisfactory patient outcomes
(Heidi Hopkins, unpublished data). It remains to be seen whether these more intensive training
programs can be taken to national scale in endemic countries in Africa. Anecdotal evidence from
Southeast Asia suggests community workers may be more responsive to RDTs than trained
clinical staff.
12
Resources for the Future
Whitty et al.
To improve the use of diagnostic tests, we need to know a great deal more about why
clinicians prescribe in the patterns described. Investigation is only just beginning, but initial data
from Tanzania provide fairly clear indications (Chandler et al. 2008). One reason Tanzanian
clinicians frequently gave was that patients expected this overprescription, but observational and
anthropological studies demonstrate that this belief is incorrect (Chandler et al. 2008). Patients in
outpatient settings stated fairly clearly that if they wanted an antimalarial they would simply buy
one, and that they come to formal health-care settings for a diagnosis. Clinicians appeared to
make malaria treatment decisions on the basis of complex mind-lines involving a mixture of
conventional clinical logic and diagnostic algorithms on the one hand, and social factors with no
obvious basis in clinical logic on the other. They also use tests to confirm their suspicions, rather
than as a way to make a diagnosis or allocate treatment. Because malaria is the most common
diagnosis outpatient clinicians in Africa make, leading to an ingrained process reinforced every
working day, evidence from other settings (e.g., antibiotic prescribing in the UK and United
States) suggests that changing behavior will not be straightforward.
What Is Relevant for AMFm?
Are these data helpful in assessing the potential impact of RDTs in the settings where
AMFm would expand ACT access? Yes and no. The key points are that just because tests are
available and used does not mean that they will change prescribing behavior, and that the
theoretical effectiveness and cost-effectiveness of RDTs may be much higher than their actual
effectiveness and cost-effectiveness. The data cannot be used directly to predict how
shopkeepers and peripheral dispensers would use RDTs. It would not be sensible to deploy
RDTs as part of AMFm except with a strong operational research component in pilot studies to
assess the impact in practice. There is at least a possibility that shopkeepers and others may be
more likely than doctors and clinical officers to respond to the results of diagnostic tests.
The Situation in Asia
The situation on diagnosis in Asia is even more complex than in Africa. The vastly
different epidemiological settings across the continent, the co-endemicity of two species of
malaria (vivax and falciparum) in varying proportions, and the scarcity of data on which to base
policy contribute to this complexity. The number of settings where malaria endemicity is as high
as Africa is relatively small but includes Papua New Guinea and parts of Indonesia, Assam in
India, Yemen and possibly Burma. Elsewhere the incidence is generally far lower, though the
13
Resources for the Future
Whitty et al.
large total populations make the sheer numbers of malaria cases substantial and the numbers of
febrile cases in malaria-endemic areas vast.
There are concerns about the availability of ACTs in most of Asia, where access remains
low. Where they are available, ACTs are not rationally prescribed (Joshi et al. 2008).
Worryingly, recent evidence from Cambodia indicates that empiric over-the-counter treatment
with artemisinin monotherapies is common and of long standing (Yeung et al. 2008) and that this
is associated with detection of reduced parasite susceptibility to artemisinin drugs. In addition,
the level of counterfeit and substandard drugs is high, which reduces treatment effectiveness,
may propagate resistance, and undermines public confidence in the treatments. In these areas, the
attributable fraction of febrile illness due to malaria is low, and the importance of targeting ACTs
to the right cases is considerable.
In many parts of Asia, and in particular in South Asia, the dominant parasite species is
vivax malaria. The cost-effectiveness of RDTs as estimated for falciparum malaria is likely to
decrease substantially for vivax malaria, though there are fewer models for vivax. This is partly
because vivax malaria is less likely to lead to loss of life, partly because the drugs effective
against vivax are themselves much cheaper, and partly because the tests that reliably diagnose
vivax malaria are generally more expensive than those that reliably diagnose only falciparum
malaria. Sensitivity to nonfalciparum malaria remains a challenge for RDTs.
Light microscopy has an advantage when vivax and falciparum co-exist, but the
advantage is realized only with greater training for microscopists, in that they need to be able to
differentiate between the species as well as simply to diagnose malaria and then to quantify it.
One approach is to treat all malaria with ACTs but when only a small proportion of febrile cases
are malaria, and of these 90 percent are vivax malaria, this is an expensive approach. In most
areas, ACTs will provide no additional benefit over cheaper drugs in terms of curing acute vivax
malaria, reducing vivax gametocyte carriage (Kolaczinski et al. 2007), and are unlikely to
eliminate hypnozoites. Emergent drug resistance to chloroquine may change this dichotomy to a
unitary treatment with ACTs, though the switch, if it occurs, is some time away outside
Southeast Asia.
When falciparum is the major species, treatment of all malaria cases with ACT will be
more cost-effective. However, when vivax is the major species, falciparum will be treated more
frequently with (potentially ineffective) drugs unless accurate diagnosis is available.
Clearly there is a central role for diagnosis in Asia, but what form this could and should
take requires careful evaluation. Prescriber practices and the role of diagnostics in Asia receive
14
Resources for the Future
Whitty et al.
less attention than the sensitivity and specificity of RDTs (Kolaczinski et al. 2004; Bharti et al.
2008). The role of the private sector in Asia is understudied as well and at best speculative. This
lack of data, coupled with the complex array of epidemiological settings, makes it more difficult
to reach conclusions on the role of RDTs in the region. Logically, of course, RDTs do have a
role, and many governments are seeking to bring RDT use to scale despite the lack of evidence
of any advantage over microscopy or even presumptive treatment.
Our view is that further research in South Asia is necessary before RDTs are deployed in
general health services, and before any part of any AMFm is extended there. We have no way of
predicting reliably what the impact of either would be on prescribing practices, and the number
of tests that could potentially be used in this setting is vast, with consequentially significant
costs. Evidence from Southeast Asia has demonstrated the potential utility of RDTs, and of
microscopy used in the periphery, but is unlikely to be translatable into the very different settings
of AMFm in Africa.
Conclusions
1. Improving diagnosis of febrile illness so that effective antimalarials are targeted to those
with malaria should be a priority, but there are no quick fixes.
2. Microscopy remains an excellent technology, but is unlikely to be relevant to the
peripheral settings where AMFm would make a difference, although experience in Asia
suggests that microscopy can be useful at the periphery in some settings.
3. There are sensitive and specific RDTs, and these could be used in the periphery, but
technical questions over heat stability, operational safety, and financial incentives for use
are significant concerns.
4. RDTs can be cost effective over a wide range of epidemiological settings—but only if
their use leads to changes in prescribing.
5. Overdiagnosis of malaria is substantial in the formal health-care sector throughout Africa,
and very possibly in Asia, based on clinical symptoms alone.
6. Evidence that deploying RDTs with a limited training package will change the
prescribing behavior in the formal health-care sector is disappointing. Changing
prescribing behavior will be challenging. This may or may not be relevant to the sectors
where AMFm will have impact.
7. The low incidence of malaria and high proportion of vivax malaria complicates the
diagnostic picture in South Asia, and we have relatively little data on diagnostic and
prescribing practices there.
15
Resources for the Future
Whitty et al.
8. The importance of diagnostics increases as the incidence of malaria, and thus the
prevalence of malaria in febrile children, decreases. This is likely to be particularly
important for the later phases of the roll-out of AMFm.
We believe that AMFm will have to take diagnosis of malaria, and febrile illness more
generally, into account if it is to have the maximum impact and remain sustainable and
acceptably cost effective, but that this is likely to be more important in later stages as the roll-out
occurs in areas with lower transmission (or as transmission in first-phase countries drops). In
Africa, a number of actions can be taken now, and in the wings are a number of interventions
such randomised trials of community education on which data will soon be available to help
guide policy. In Asia, substantially more information is needed before any recommendations can
be made.
What would need to be undertaken depends on the setting, which we divide broadly into
formal health-care outpatient settings where microscopy is available; formal health-care settings
where microscopy is not available (most clinics in Africa); the private sector, subdivided into the
formal health-care sector and small unregulated shops; and community drug distributors, such as
in home-based management of fever (HBMF) programs.
Within the formal health-care sector, in hospital settings where microscopy is available,
the priorities are clearest because the evidence base is best. The need to increase the proportion
of febrile children and adults tested for malaria, and then to reduce the overprescription of
antimalarials to those with negative tests, is clear. This can only happen, however, when
diagnostic facilities have been improved. This may be achieved by maintaining microscopy
standards and materials or, if this proves impractical, by supplying RDTs. The key, however, is
not in technology, but in attitude and behavior change amongst clinical prescribers. The formal
health-care sector is however not one likely to be important for AMFm; the Global Fund already
provides subsidized ACTs in this setting to most endemic countries.
In the formal sector where microscopy is not available, such as in clinics, it is clear that
some diagnostic facilities would be an improvement on no diagnostic facilities and that
deploying RDTs is likely to be the only practical solution in most settings. Ongoing research is
needed to assess the impact of RDT implementation on actual prescribing behavior and, in all
probability, complex behavioral interventions as well.
Using RDTs in the formal private sector seems an entirely rational choice, but a relatively
small proportion of the target beneficiaries of AMFm also use the formal private sector. Those
16
Resources for the Future
Whitty et al.
who do are at relatively lower risk of malaria than other groups given that, in most African and
Asian countries, malaria is primarily a disease of the poor and of those in rural communities.
The largest unknown area, which offers the greatest potential gains, is the informal
private sector, such as general shops, specialized pharmacies, and chemical sellers. Users of this
sector are often the poorest and the most likely to be affected positively by the AMFm. Massive
overprescription of antimalarials in this sector (which has long been tolerated with cheaper
antimalarials) could significantly reduce sustainability of the AMFm system. We do not yet have
information about how RDTs would be used in this sector in Africa, though some studies are
under way. It is essential that deploying RDTs neither decreases the number of people with true
malaria treated with an antimalarial (e.g., false-negative results), nor increases the proportion of
people treated for malaria who do not have it (meaning they might die of other causes),
significantly increases the risk of transmission of blood-borne viruses. This last element is
particularly important. It would be a serious setback from a public health point of view if, in
promoting better diagnosis of malaria, we also promote increased transmission of HIV or
hepatitis B. This is not a trivial risk: shops working on small profit margins may try to make a
little more money by re-using lancets. The safety of the shopkeepers should also be considered,
given the small but real risk of needle-stick injuries. How to provide proper incineration of shopused lancets would be another important operational question. Changing shopkeeper behavior is
possible but requires intensive piloting and adapting to local conditions (Marsh et al. 1999). In
our view, therefore, deploying RDTs in this setting without thoughtful operational assessment of
safety and effectiveness would be premature and potentially even dangerous.
In considering RDT use by community volunteers, such as in HBMF programs, the same
concerns apply. Some promising early work has been done to evaluate the implementation of
RDTs in Zambia’s HBMF program, and evidence should be forthcoming within the next year or
two (Harvey et al. 2008). It remains to be demonstrated that successes with RDTs in Southeast
Asia in this setting can be replicated in Africa.
In the longer term, and as malaria decreases in some settings, leading to a reduction in the
proportion of children with febrile illness presenting with malaria, we may need to consider, at
least in some settings, syndromic management of fever with a combination of antimalarials and
antibiotics. Although clinicians already use syndromic treatment to some extent in individual
cases where the diagnosis is in doubt, deploying it on policy to all febrile children is some way in
the future and would meet substantial resistance. Concerns about drug safety and the promotion
of antibiotic resistance are legitimate. For the moment, however, the choices are between
improving and not improving diagnosis, and in that choice the overall goal must be clear. We
17
Resources for the Future
Whitty et al.
now have the technology to make a proper malaria diagnosis, and though technology is
continuing to improve, the biggest stumbling block continues to be behavioral change rather than
technological problems.
The AMFm provides a remarkable opportunity to improve the whole management of
febrile illness in Africa. This will require operational research alongside deployment but, if
undertaken properly, could have positive ramifications for all the other causes of febrile illness
and thus reduce mortality beyond the malaria-attributable fraction that the AMFm is addressing
directly.
18
Resources for the Future
Whitty et al.
References
Barnes, K.I., D.N. Durrheim, F. Little, A. Jackson, U. Mehta, E. Allen, S.S. Dlamini, J. Tsoka,
B. Bredenkamp, D.J. Mthembu, N.J. White, and B.L. Sharp. 2005. Effect of ArtemetherLumefantrine Policy and Improved Vector Control on Malaria Burden in Kwazulu-Natal,
South Africa. PLoS Medicine 2(11): e330.
Berkley, J.A., K. Maitland, I. Mwangi, C. Ngetsa, S. Mwarumba, B.S. Lowe, C.R. Newton, K.
Marsh, J.A. Scott, and M. English. 2005. Use of Clinical Syndromes to Target Antibiotic
Prescribing in Seriously Ill Children in Malaria Endemic Area: Observational Study.
British Medical Journal 330(7498): 995.
Berkley, J.A., I. Mwangi, F. Mellington, S. Mwarumba, and K. Marsh. 1999. Cerebral Malaria
Versus Bacterial Meningitis in Children with Impaired Consciousness. QJM 92(3): 1517.
Bharti P.K., N. Silawat, P.P. Singh, et al. 2008. The Usefulness of a New Rapid Diagnostic Test,
the First Response Malaria Combo (pLDH/HRP2) Card Test, for Malaria Diagnosis in
the Forested Belt of Central India. Malaria Journal. 11 (7):126.
Bhattarai, A., A.S. Ali, S.P. Kachur, A. Martensson, A.K. Abbas, R. Khatib, A.W. Al-Mafazy,
M. Ramsan, G. Rotllant, J.F. Gerstenmaier, F. Molteni, S. Abdulla, S.M. Montgomery,
A. Kaneko, and A. Bjorkman. 2007. Impact of Artemisinin-Based Combination Therapy
and Insecticide-Treated Nets on Malaria Burden in Zanzibar. PLoS Medicine 4(11): e309.
Breman, J.G. 2001. The Ears of the Hippopotamus: Manifestations, Determinants, and Estimates
of the Malaria Burden. American Journal of Tropical Medicine and Hygiene 64(1-2
Supplement): 1-11.
Bualombai, P., S. Prajakwong, N. Aussawatheerakul, K. Congpoung, S. Sudathip, K. Thimasarn,
J. Sirichaisinthop, K. Indaratna, C. Kidson, and M. Srisuphanand. 2003. Determining
Cost-Effectiveness and Cost Component of Three Malaria Diagnostic Models Being
Used in Remote Non-Microscope Areas. Southeast Asian Journal of Tropical Medicine
and Public Health 34(2): 322-33.
Chandler, C.I., S. Chonya, G. Boniface, K. Juma, H. Reyburn, and C.J. Whitty. 2008. The
Importance of Context in Malaria Diagnosis and Treatment Decisions - a Quantitative
Analysis of Observed Clinical Encounters in Tanzania. Tropical Medicine and
International Health (Epub ahead of print).
19
Resources for the Future
Whitty et al.
Chandler, C.I., C. Jones, G. Boniface, K. Juma, H. Reyburn, and C.J. Whitty. 2008. Guidelines
and Mindlines: Why Do Clinical Staff over-Diagnose Malaria in Tanzania? A Qualitative
Study. Malaria Journal 7: 53.
Chandler, C.I., R. Mwangi, H. Mbakilwa, R. Olomi, C.J. Whitty, and H. Reyburn. 2008. Malaria
Overdiagnosis: Is Patient Pressure the Problem? Health Policy and Planning 23(3): 1708.
Chandramohan, D., S. Jaffar, and B. Greenwood. 2002. Use of Clinical Algorithms for
Diagnosing Malaria. Tropical Medicine and International Health 7(1): 45-52.
Chiodini, P.L., K. Bowers, P. Jorgensen, J.W. Barnwell, K.K. Grady, J. Luchavez, A.H. Moody,
A. Cenizal, and D. Bell. 2007. The Heat Stability of Plasmodium Lactate
Dehydrogenase-Based and Histidine-Rich Protein 2-Based Malaria Rapid Diagnostic
Tests. Transactions of the Royal Society of Tropical Medicine and Hygiene 101(4): 3317.
Coleman, R.E., N. Maneechai, N. Rachaphaew, C. Kumpitak, R.S. Miller, V. Soyseng, K.
Thimasarn, and J. Sattabongkot. 2002. Comparison of Field and Expert Laboratory
Microscopy for Active Surveillance for Asymptomatic Plasmodium Falciparum and
Plasmodium Vivax in Western Thailand. American Journal of Tropical Medicine and
Hygiene 67(2): 141-4.
Craig, M.H., B.L. Bredenkamp, C.H. Williams, E.J. Rossouw, V.J. Kelly, I. Kleinschmidt, A.
Martineau, and G.F. Henry. 2002. Field and Laboratory Comparative Evaluation of Ten
Rapid Malaria Diagnostic Tests. Transactions of the Royal Society of Tropical Medicine
and Hygiene 96(3): 258-65.
Duong, S., P. Lim, T. Fandeur, R. Tsuyuoka, and C. Wongsrichanalai. 2004. Importance of
Protection of Antimalarial Combination Therapies. Lancet 364(9447): 1754-5.
Durrheim, D.N., P.J. Becker, and K. Billinghurst. 1997. Diagnostic Disagreement--the Lessons
Learnt from Malaria Diagnosis in Mpumalanga. South African Medical Journal 87(8):
1016.
Gay, F., B. Traore, J. Zanoni, M. Danis, and A. Fribourg-Blanc. 1996. Direct Acridine Orange
Fluorescence Examination of Blood Slides Compared to Current Techniques for Malaria
Diagnosis. Transactions of the Royal Society of Tropical Medicine and Hygiene 90(5):
516-8.
20
Resources for the Future
Whitty et al.
Gray, A.D., M.T. Akin, R. McLean, and C.R. Valeri. 1991. Evaluation of the Quantitative Buffy
Coat (Qbc) Method to Detect Malaria-Infected Red Blood Cells. Military Medicine
156(5): 241-5.
Guerra, C.A., P.W. Gikandi, A.J. Tatem, A.M. Noor, D.L. Smith, S.I. Hay, and R.W. Snow.
2008. The Limits and Intensity of Plasmodium Falciparum Transmission: Implications
for Malaria Control and Elimination Worldwide. PLoS Medicine 5(2): e38.
Guthmann, J.P., A. Ruiz, G. Priotto, J. Kiguli, L. Bonte, and D. Legros. 2002. Validity,
Reliability and Ease of Use in the Field of Five Rapid Tests for the Diagnosis of
Plasmodium Falciparum Malaria in Uganda. Transactions of the Royal Society of
Tropical Medicine and Hygiene 96(3): 254-7.
Guy, R., P. Liu, P. Pennefather, and I. Crandall. 2007. The Use of Fluorescence Enhancement to
Improve the Microscopic Diagnosis of Falciparum Malaria. Malaria Journal 6: 89.
Hamer, D.H., M. Ndhlovu, D. Zurovac, M. Fox, K. Yeboah-Antwi, P. Chanda, N. Sipilinyambe,
J.L. Simon, and R.W. Snow. 2007. Improved Diagnostic Testing and Malaria Treatment
Practices in Zambia. Journal of the American Medical Association 297(20): 2227-2231.
Harvey, S.A., L. Jennings, M. Chinyama, F. Masaninga, K. Mulholland, and D.R. Bell. 2008.
Improving Community Health Worker Use of Malaria Rapid Diagnostic Tests in Zambia:
Package Instructions, Job Aid and Job Aid-Plus-Training. Malaria Journal 7(1): 160.
Hopkins, H., W. Kambale, M.R. Kamya, S.G. Staedke, G. Dorsey, and P.J. Rosenthal. 2007.
Comparison of Hrp2- and Pldh-Based Rapid Diagnostic Tests for Malaria with
Longitudinal Follow-up in Kampala, Uganda. American Journal of Tropical Medicine
and Hygiene 76(6): 1092-1097.
Jonkman, A., R.A. Chibwe, C.O. Khoromana, U.L. Liabunya, M.E. Chaponda, G.E. Kandiero,
M.E. Molyneux, and T.E. Taylor. 1995. Cost-Saving through Microscopy-Based Versus
Presumptive Diagnosis of Malaria in Adult Outpatients in Malawi. Bulletin of the World
Health Organization 73(2): 223-7.
Jorgensen, P., L. Chanthap, A. Rebueno, R. Tsuyuoka, and D. Bell. 2006. Malaria Rapid
Diagnostic Tests in Tropical Climates: The Need for a Cool Chain. American Journal of
Tropical Medicine and Hygiene 74(5): 750-4.
Joshi R., J.M. Colford Jr., A.L. Reingold, and S. Kalantri. Nonmalarial Acute Undifferentiated
Fever in a Rural Hospital in Central India: Diagnostic Uncertainty and Overtreatment
21
Resources for the Future
Whitty et al.
with Antimalarial Agents. 2008. American Journal of Tropical Medicine and Hygiene.
78(3):393-9.
Kallander, K., H. Hildenwall, P. Waiswa, E. Galiwango, S. Peterson, and G. Pariyo. 2008.
Delayed Care Seeking for Fatal Pneumonia in Children Aged under Five Years in
Uganda: A Case-Series Study. Bulletin of the World Health Organization 86(5): 332-8.
Kolaczinski J.,N. Mohammed, I. Ali , M. Ali, N. Khan, N. Ezard, and M.Rowland. 2004.
Comparison of the OptiMAL Rapid Antigen Test with Field Microscopy for the
Detection of Plasmodium vivax and P. falciparum: Considerations for the Application of
the Rapid Test in Afghanistan. Annals of Tropical Medicine and Parasitology. 98(1):1520.
Kolaczinski K., N. Durrani, S. Rahim, and M. Rowland. Sulfadoxine-Pyrimethamine Plus
Artesunate Compared with Chloroquine for the Treatment of vivax Malaria in Areas Coendemic for Plasmodium falciparum and P. vivax: A Randomised Non-inferiority Trial in
Eastern Afghanistan. 2007. Transactions of the Royal Society of Tropical Medicine and
Hygiene. 101(11):1081-1087.
Lubell, Y., H. Hopkins, C.J. Whitty, S.G. Staedke, and A. Mills. 2008. An Interactive Model for
the Assessment of the Economic Costs and Benefits of Different Rapid Diagnostic Tests
for Malaria. Malaria Journal 7(1): 21.
Lubell, Y., H. Reyburn, H. Mbakilwa, R. Mwangi, K. Chonya, C.J. Whitty, and A. Mills. 2007.
The Cost-Effectiveness of Parasitologic Diagnosis for Malaria-Suspected Patients in an
Era of Combination Therapy. American Journal of Tropical Medicine and Hygiene 77(6
Supplement): 128-32.
Lubell, Y., H. Reyburn, H. Mbakilwa, R. Mwangi, S. Chonya, C.J. Whitty, and A. Mills. 2008.
The Impact of Response to the Results of Diagnostic Tests for Malaria: Cost-Benefit
Analysis. British Medical Journal 336(7637): 202-5.
Luxemburger, C., F. Nosten, D.E. Kyle, L. Kiricharoen, T. Chongsuphajaisiddhi, and N.J. White.
1998. Clinical Features Cannot Predict a Diagnosis of Malaria or Differentiate the
Infecting Species in Children Living in an Area of Low Transmission. Transactions of
the Royal Society of Tropical Medicine and Hygiene 92(1): 45-9.
Maiteki-Sebuguzi, C., P. Jagannathan, V.M. Yau, T.D. Clark, D. Njama-Meya, B. Nzarubara,
A.O. Talisuna, et al. 2008. Safety and Tolerability of Combination Antimalarial
22
Resources for the Future
Whitty et al.
Therapies for Uncomplicated Falciparum Malaria in Ugandan Children. Malaria Journal
7: 106.
Marsh V.M., W.M. Mutemi, J. Muturi, A. Haaland, W.M. Watkins, G. Otieno, and K. Marsh.
1999. Changing Home Treatment of Childhood Fevers by Training Shop Keepers in
Rural Kenya. Tropical Medicine and International Health. 4(5):383-9.
Mayxay, M., P.N. Newton, S. Yeung, T. Pongvongsa, S. Phompida, R. Phetsouvanh, and N.J.
White. 2004. Short Communication: An Assessment of the Use of Malaria Rapid Tests
by Village Health Volunteers in Rural Laos. Tropical Medicine and International Health
9(3): 325-9.
Mayxay, M., S. Pukrittayakamee, K. Chotivanich, S. Looareesuwan, and N.J. White. 2001.
Persistence of Plasmodium Falciparum Hrp-2 in Successfully Treated Acute Falciparum
Malaria. Transactions of the Royal Society of Tropical Medicine and Hygiene 95(2): 17982.
Mockenhaupt, F.P., G. Bedu-Addo, C. von Gaertner, R. Boye, K. Fricke, I. Hannibal, F.
Karakaya, et al. 2006. Detection and Clinical Manifestation of Placental Malaria in
Southern Ghana. Malari.a Journal 5: 119.
Moody, A. 2002. Rapid Diagnostic Tests for Malaria Parasites. Clinical Microbiology Reviews
15(1): 66-78.
Ndyomugyenyi, R., P. Magnussen, and S. Clarke. 2007. Diagnosis and Treatment of Malaria in
Peripheral Health Facilities in Uganda: Findings from an Area of Low Transmission in
South-Western Uganda. Malaria Journal 6: 39.
Nyarango, P.M., T. Gebremeskel, G. Mebrahtu, J. Mufunda, U. Abdulmumini, A. Ogbamariam,
A. Kosia, et al. 2006. A Steep Decline of Malaria Morbidity and Mortality Trends in
Eritrea between 2000 and 2004: The Effect of Combination of Control Methods. Malaria
Journal 5: 33.
Okiro, E.A., S.I. Hay, P.W. Gikandi, S.K. Sharif, A.M. Noor, N. Peshu, K. Marsh, and R.W.
Snow. 2007. The Decline in Paediatric Malaria Admissions on the Coast of Kenya.
Malaria Journal 6: 151.
Olivar, M., M. Develoux, A. Chegou Abari, and L. Loutan. 1991. Presumptive Diagnosis of
Malaria Results in a Significant Risk of Mistreatment of Children in Urban Sahel.
Transactions of the Royal Society of Tropical Medicine and Hygiene 85(6): 729-30.
23
Resources for the Future
Whitty et al.
Orimadegun, A.E., O.K. Amodu, P.E. Olumese, and O.O. Omotade. 2008. Early Home
Treatment of Childhood Fevers with Ineffective Antimalarials Is Deleterious in the
Outcome of Severe Malaria. Malaria Journal 7: 143.
Payne, D. 1988. Did Medicated Salt Hasten the Spread of Chloroquine Resistance in
Plasmodium Falciparum? Parasitology Today 4(4): 112-5.
Premji, Z., J.N. Minjas, and C.J. Shiff. 1994. Laboratory Diagnosis of Malaria by Village Health
Workers Using the Rapid Manual Parasight-F Test. Transactions of the Royal Society of
Tropical Medicine and Hygiene 88(4): 418.
Redd, S.C., P.N. Kazembe, S.P. Luby, O. Nwanyanwu, A.W. Hightower, C. Ziba, J.J. Wirima,
L. Chitsulo, C. Franco, and M. Olivar. 1996. Clinical Algorithm for Treatment of
Plasmodium Falciparum Malaria in Children. Lancet 347(8996): 223-7.
Rennie, W., R. Phetsouvanh, S. Lupisan, V. Vanisaveth, B. Hongvanthong, S. Phompida, P.
Alday, et al. 2007. Minimising Human Error in Malaria Rapid Diagnosis: Clarity of
Written Instructions and Health Worker Performance. Transactions of the Royal Society
of Tropical Medicine and Hygiene 101(1): 9-18.
Reyburn, H., H. Mbakilwa, R. Mwangi, O. Mwerinde, R. Olomi, C. Drakeley, and C.J. Whitty.
2007. Rapid Diagnostic Tests Compared with Malaria Microscopy for Guiding
Outpatient Treatment of Febrile Illness in Tanzania: Randomised Trial. British Medical
Journal 334(7590): 403.
Reyburn, H., R. Mbatia, C. Drakeley, I. Carneiro, E. Mwakasungula, O. Mwerinde, K. Saganda,
et al. 2004. Overdiagnosis of Malaria in Patients with Severe Febrile Illness in Tanzania:
A Prospective Study. British Medical Journal 329(7476): 1212.
Shillcutt, S., C. Morel, C. Goodman, P. Coleman, D. Bell, C.J. Whitty, and A. Mills. 2008. CostEffectiveness of Malaria Diagnostic Methods in Sub-Saharan Africa in an Era of
Combination Therapy. Bulletin of the World Health Organization 86(2): 101-10.
Sievers, A.C., J. Lewey, P. Musafiri, M.F. Franke, B.J. Bucyibaruta, S.N. Stulac, M.L. Rich, C.
Karema, and J.P. Daily. 2008. Reduced Paediatric Hospitalizations for Malaria and
Febrile Illness Patterns Following Implementation of Community-Based Malaria Control
Programme in Rural Rwanda. Malaria Journal 7(1): 167.
Singh, N., and M.M. Shukla. 2002. Short Report: Field Evaluation of Posttreatment Sensitivity
for Monitoring Parasite Clearance of Plasmodium Falciparum Malaria by Use of the
24
Resources for the Future
Whitty et al.
Determine Malaria Pf Test in Central India. American Journal of Tropical Medicine and
Hygiene 66(3): 314-6.
Skarbinski J., P. Ouma, L. Causer, S. Kariuki, J.Barnwell, J.Alaii, A.M. de Oliveira, et al. 2007.
Introduction of Malaria Rapid Diagnostic Tests, New Guidelines, and ArtemetherLumefantrine in Kenya: A Cluster Randomized Trial. Presented at the American Society
of Tropical Medicine and Hygiene 56th Annual Meeting. November, 2007, Philadelphia,
PA.
Snow, R.W., E. Eckert, and A. Teklehaimanot. 2003. Estimating the Needs for Artesunate-Based
Combination Therapy for Malaria Case-Management in Africa. Trends in Parasitology
19(8): 363-9.
Sowunmi, A., and J.A. Akindele. 1993. Presumptive Diagnosis of Malaria in Infants in an
Endemic Area. Transactions of the Royal Society of Tropical Medicine and Hygiene
87(4): 422.
Ssekabira, U., Bukirwa H, Hopkins H, Namagembe A, Weaver MR, Mpanga Sebuyira L, Quick
L, et al. submitted for publication. Improved Malaria Case Management Following
Integrated Team-Based Training of Health Care Workers in Uganda.
Stow, N.W., J.K. Torrens, and J. Walker. 1999. An Assessment of the Accuracy of Clinical
Diagnosis, Local Microscopy and a Rapid Immunochromatographic Card Test in
Comparison with Expert Microscopy in the Diagnosis of Malaria in Rural Kenya.
Transactions of the Royal Society of Tropical Medicine and Hygiene 93(5): 519-20.
Swarthout, T.D., H. Counihan, R.K. Senga, and I. van den Broek. 2007. Paracheck-Pf(R)
Accuracy and Recently Treated Plasmodium Falciparum Infections: Is There a Risk of
over-Diagnosis? Malaria Journal 6(1): 58.
Talisuna, A.O., P. Bloland, and U. D'Alessandro. 2004. History, Dynamics, and Public Health
Importance of Malaria Parasite Resistance. Clinical Microbiology Reviews 17(1): 235-54.
Tarimo, D.S., J.N. Minjas, and I.C. Bygbjerg. 2001. Malaria Diagnosis and Treatment under the
Strategy of the Integrated Management of Childhood Illness (Imci): Relevance of
Laboratory Support from the Rapid Immunochromatographic Tests of Ict Malaria
P.F/P.V and Optimal. Annals of Tropical Medicine and Parasitology 95(5): 437-44.
Tjitra, E., S. Suprianto, M.E. Dyer, B.J. Currie, and N.M. Anstey. 2001. Detection of Histidine
Rich Protein 2 and Panmalarial Ict Malaria Pf/Pv Test Antigens after Chloroquine
Treatment of Uncomplicated Falciparum Malaria Does Not Reliably Predict Treatment
25
Resources for the Future
Whitty et al.
Outcome in Eastern Indonesia. American Journal of Tropical Medicine and Hygiene
65(5): 593-8.
Uneke, C.J. 2008. Diagnosis of Plasmodium Falciparum Malaria in Pregnancy in Sub-Saharan
Africa: The Challenges and Public Health Implications. Parasitology Research 102(3):
333-42.
White, N.J., and P.L. Olliaro. 1996. Strategies for the Prevention of Antimalarial Drug
Resistance: Rationale for Combination Chemotherapy for Malaria. Parasitology Today
12(10): 399-401.
WHO (World Health Organization).2006. Guidelines for the Treatment of Malaria. Geneva,
Switzerland: WHO.
WHO (World Health Organization). 2008. Global Malaria Control and Elimination: Report of a
Meeting on Containment of Artemisinin Tolerance. Geneva, Switzerland: WHO.
WHO/WPRO (World Health Organization/Western Pacific Regional Office). 2008. List of
Known Commercially-Available Antigen-Detecting Malaria Rdts.
http://www.wpro.who.int/NR/rdonlyres/9493AF24-5BAC-4CC9-ACF26E1B7EED1E85/0/MD_table32_ISO131485criteriarev240808.xls. (accessed September
16, 2008).
Yeung, S., W. Van Damme, D. Socheat, N.J. White, and A. Mills. 2008. Cost of Increasing
Access to Artemisinin Combination Therapy: The Cambodian Experience. Malaria
Journal 7: 84.
Yeung S., W. Van Damme, D. Socheat, N.J. White, and A. Mills. Access to Artemisinin
Combination Therapy for Malaria in Remote Areas of Cambodia. 2008. Malaria Journal.
29 (7):96.
Zurovac, D., J. Njogu, W. Akhwale, D.H. Hamer, B.A. Larson, and R.W. Snow. 2008. Effects of
Revised Diagnostic Recommendations on Malaria Treatment Practices across Age
Groups in Kenya.
26
Resources for the Future
Whitty et al.
Table 1. Advantages and disadvantages of different diagnostic test methods
Diagnostic Method Advantages Disadvantages • Provides additional diagnostic information: quantification of parasite density, identification of parasite species, assessment • Reliable results require well‐maintained microscopes, reagents and supplies, as well as electricity (or strong of hematological abnormalities, identification of certain other sunlight) infectious agents if present Light microscopy Rapid diagnostic tests (RDTs) • Useful for monitoring response to treatment • Need for skilled and motivated laboratory personnel • If quality maintained, likely cost advantage over other diagnostic methods (RDTs and empiric treatment) where patient volume is high • More time‐consuming and labor‐intensive than other diagnostic methods (RDTs and empiric treatment) • Relatively simple to use, can be performed by health workers with limited formal training • Give only “yes or no” information on parasitemia; not quantitative • Do not require special equipment or electricity • Susceptible to degradation and loss of sensitivity when exposed to high temperatures and humidity • Likely cost advantage over other diagnostic methods (microscopy and empiric treatment) where patient volume is •Require fingerstick for blood, with some risk of transmission of bloodborne diseases, including HIV and hepatitis B relatively low RDT: HRP2 • Sensitivity maintained at lower parasite densities • Detects P. falciparum only (some tests combine HRP2 detection with pLDH or aldolase detection to allow diagnosis (histidine‐rich protein 2) • Relatively stable in typical storage conditions in endemic areas of non‐falciparum species as well) (though may vary by manufacturer, etc) • Antigenemia persists post‐treatment, which precludes use to monitor treatment response, and may lead to false‐positive results in areas of intense transmission RDT: pLDH • Detects P. falciparum, P. vivax, P. ovale, and P. malariae (some pLDH‐based tests can distinguish P. falciparum and P. vivax from other species) (plasmodium lactate dehydrogenase) • Consistently mirrors parasitemia, so can be used to monitor treatment response 28
• Sensitivity drops at lower parasite densities • Less stable at typical storage conditions in endemic areas Resources for the Future
28